SVM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import sys\n",
    "import numpy as np\n",
    "import random\n",
    "import math\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def appenddata(d):\n",
    "    x = d[0]\n",
    "    y = d[1]\n",
    "    d.append(x*x)\n",
    "    d.append(y*y)\n",
    "    d.append(x*y)\n",
    "    d.append(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x74b3358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "points = [\n",
    "    [2, 5],\n",
    "    [5, 1],\n",
    "    [7, 8],\n",
    "    [1, -5],\n",
    "    [-3, 5],\n",
    "    [6, -2],\n",
    "    [-1, 0],\n",
    "    [2, 8],\n",
    "    [3, 10],\n",
    "    [4, 7],\n",
    "    [0, 8],\n",
    "    [1, 4],\n",
    "    [4,-4]\n",
    "]\n",
    "\n",
    "for p in points:\n",
    "    appenddata(p)\n",
    "\n",
    "X = np.array(points)\n",
    "\n",
    "y = np.array([1,1,-1,-1,-1,-1,-1,1,1,1,1,-1,-1])\n",
    "\n",
    "for d, sample in enumerate(X):\n",
    "    if y[d] == -1:\n",
    "        plt.scatter(sample[0], sample[1], s=120, marker=\".\", linewidths=2,c=\"blue\")\n",
    "    else:\n",
    "        plt.scatter(sample[0], sample[1], s=120, marker=\".\", linewidths=2,c=\"red\")        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def update_weights(X,Y,learning_rate,epochs):\n",
    "    w = np.random.rand(len(X[0]))\n",
    "    \n",
    "    for epoch in range(1, epochs):\n",
    "        for i, x in enumerate(X):\n",
    "            if Y[i] * np.dot(X[i], w) < 1:\n",
    "                w = w + learning_rate * X[i] * Y[i] + (-2  * (1 / epoch) * w)\n",
    "            else:\n",
    "                w = w + learning_rate * (-2 * (1 / epoch) * w)\n",
    "        if epoch % 5000 == 0:\n",
    "            sys.stdout.write(\"\\r{:.2f} %\".format(100*epoch/epochs))\n",
    "            sys.stdout.flush()\n",
    "            learning_rate *=0.99\n",
    "            \n",
    "            \n",
    "    sys.stdout.write(\"\\r100.00 %\")\n",
    "    sys.stdout.flush()\n",
    "    return w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100.00 %"
     ]
    }
   ],
   "source": [
    "w = update_weights(X, y, 1, 1000000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nOUMTuxBC5AxN7EIIkTM0sQshRM74/3/7l26/Xc4YAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x719ab38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = np.linspace(-5,10,75)\n",
    "xx = np.linspace(-5,10,75)\n",
    "points = []\n",
    "for i in x:\n",
    "    for j in xx:\n",
    "        points.append([i,j])\n",
    "for p in points:\n",
    "    appenddata(p)\n",
    "for x in points:    \n",
    "    label = np.dot(np.array(x), w)\n",
    "    if label < 1:\n",
    "        plt.scatter(x[0], x[1], s=20, marker=\".\", linewidth=2, c=\"r\")\n",
    "        \n",
    "\n",
    "for d, sample in enumerate(X):\n",
    "    if y[d] == -1:\n",
    "        plt.scatter(sample[0], sample[1], s=50, marker=\".\", linewidth=2, c=\"b\")\n",
    "    else:\n",
    "        plt.scatter(sample[0], sample[1], s=50, marker=\".\", linewidth=2, c=\"g\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}